Fecal steroid extraction

Fecal samples were collected at least three times per week and stored at −20°C until processing prior to hormone analysis. Samples were lyophilized, pulverized and sifted through a 0.045-inch mesh screen to remove vegetation and debris. For all samples up to 5 May 2017, 0.1 g of sifted fecal material in a 16 × 100 mm glass tube was combined with 5 ml of 90% aqueous ethanol (EMD Millipore, Billerica, MA) and boiled at 80°C for 20 min. Samples were centrifuged (Thermo Scientific Sorvall Lengend XTR, Waltham, MA) for 10 min at 1000 x g at room temperature and the supernatant was recovered. Remaining fecal material was mixed with 5 ml 90% aqueous ethanol, pulse vortexed and centrifugation was repeated. Supernatants were combined, dried under air and resuspended in 1 ml ethanol.

For samples after 5 May 2017 0.2 g sifted fecal material in a 50 ml polypropylene tube was combined with 20 ml of 80% methanol (Fisher, Waltham, MA) in sterile water and vortexed at room temperature for 30 min (Fisher Scientific MultiTube Vortexer, Waltham, MA). Samples were centrifuged for 10 min at 4000 x g at room temperature (Thermo Scientific Sorvall Lengend XTR, Waltham, MA) and the supernatant was recovered.

Transition to a new fecal extraction method was initiated to decrease sample processing time and expedite analysis. The 80:20 methanol method was chosen based on reports by Palme et al. (1997, 2013) demonstrating this as the most effective method of hormone extraction. To ensure values were comparable between methods, extractions and radioimmunoassays (RIAs) for progestogens were performed with each method simultaneously (n = 196, distributed throughout 12 extractions and assays). Progestagen values did not differ between the two extraction methods (Pearson correlation r = 0.948, Paired T-test for significant difference P = 0.735). Methanol extraction efficiency measuring recovery of added tritiated progesterone (Perkin Elmer, Waltham, MA) was 82% ± 2.8%.